# When do I do motion planning?

I am studying mobile robots, and I am focusing my attention to the unycicle. I have a doubt concerning the motion planning part. In particular I don't understand when the motion planning is done and where to put a motion planning module in a block scheme.

Suppose I want to build an artificial potential field to drive the robot to an initial position to a final position. I don't understand if I have to do first the localization, for example using an Extended Kalman Filter, or the localization comes after.

For what I have undeerstood, I first need to stabilize the system, for example using feedback linearization, and then pass the velocity inputs to the robot. Then I have seen that we do localization, which if we look at a block scheme will be in the feedback part.

But, in this block scheme I don't undesrtand where to do motion planning.

[EDIT] I have to fill my block scheme with three elemnts:

-A robot-independent module that uses the environment geometry and the assigned goal to build an artificial force field $$f(p)$$.

-A module that transforms the artificial command f(p) into actual velocity inputs $$v$$,$$ω$$.

-. A module that computes the robot state needed by the first two.

the robot I am considering is a unycicle were $$q = (p, θ)$$ is the robot configuration, with $$p = (x, y)$$, and by $$v$$,$$ω$$ its velocity inputs. An idea that came into my mind is to wrrite a block scheme where I perform input-output linearization, so I have to add a decoupling matrix $$T(x)$$, which will work as module that transforms the artificial command f(p) into the velocities, but the rest I don't know much how to do. For example what I tried is: but for sure it is not correct, since I am not considering $$f(q)$$ and I don't know how to place the motion planning part.

I am considering artificial potential field.

Can somebody help me?

[EDIT] ingnore the arrow going a little bit out in the feedback, it is a mistake I did when using the software. Sorry it is the first time I do i blck diagram on a software.

• Thanks for your question J.D., but I'm afraid that it is not clear what you are asking. We prefer practical, answerable questions based on actual problems that you face, so it's a good idea to include details of what you want to understand, what you've researched so far, what you found & what you expected to find. Perhaps adding the block diagrams you already haver to explain what you have so far would help. If you can edit your question to make it clearer we can re-open it for you. – Mark Booth Jan 6 at 15:30
• Motion planning would typically do the job that your artificial potential fields do. In your block diagram, it would generate intermediate x,y goals on the way to the final goal pose. Normally motion planning happens in parallel with your controls loop (in order to respond to tracking error, dynamic environments, etc). – combo Jan 15 at 17:39